Collaborative configurable systems are computer software systems, computer hardware systems or combinations thereof, wherein a system behavior is controlled and modifiable by a set of configuration data. Examples of such systems include, (but are not limited to), database management systems, electronic messaging and mail software and servers, electronic commerce systems, computer operating systems, (especially multi user system such as Unix® or Microsoft Windows NT®), database management systems, configurable network equipment such as routers switches, and the like. Those systems may comprise a single computer or multiple, interconnected computers, known as multi-server systems.
Configuration data is a collection of configuration parameters and data objects that control and modify at least a portion of the configurable system behavior. Complex configurable systems typically have a large number of such configuration parameters, often ranging in the hundreds and sometimes thousands of parameters. Often cross-dependencies are observed, where a change of one parameter will cause unforeseen and unintended system behavior requiring change to other pararmeters. Configuring such systems is difficult, and requires a high level of knowledge, skill and experience. The problem becomes exponentially more complex when several such systems are connected together, for example a number of cooperating computer servers, distributed e-mail servers, distributed database systems, and cooperating network switches and routers or any other distributed configurable system. In those cases, changing a parameter on one system may even disrupt the operation of another.
Due to the complexity involved in the configuration, once configured, the configurable system becomes a system of its own merit, a special case separate from the general case of the generic non-configured system. While the configured system behavior may be constructed and understood using documentation for the generic system combined with configuration data, such construction is time consuming and often wasteful and expensive. Therefore it is highly desirable to document the specific case of every instance of the configured system separately, in order to ease maintenance and troubleshooting, as well as to facilitate knowledge transfer to new personnel, and generally increase operational efficiency.
Generating such documentation is an extremely tedious job that often requires a highly trained professional to perform. Since the configuration of such systems changes with time, the task is a continual one, taxing information technology personnel. And thus automatic generation of such documentation as shown by the present invention is clearly advantageous. In these specifications, the word “automatic” and its derivative means an operation that is, or may be, created or coordinated primarily by a machine or a computer, especially as compared to chores and tasks that were formerly done primarily by human labor. In an automatic documentation generation step for example, little or no manual user intervention is necessary. Manual intervention implies steps such as manual data entry of configuration parameters, repeated text entries, etc.
Constraint programming (alternatively referred to as “constraints-based reasoning”, or “constraint satisfaction”) is a technique, most appropriate for computer use, which is based on creating a model of a problem in terms of the requirements for a solution. By defining acceptable values to variables and constraints that define and optionally quantify the allowed relationships among the values assigned to variables, a formal representation of the problem is produced. Standard constraints programming methods can use this representation to find a solution to the problem. The technique also offers an efficient approach to problem solving by making inferences on possible solutions. The technology is well known in the art. By way of example, constraints programming was used as an aid to system configuration as described in U.S. Pat. No. 5,708,798 to Lynch et al. and in PCT/US97/21218 to Elfe et al. Constraint programming easily lends itself to diagnostics and also to the automatic documentation process described herewith.
While most configurable systems provide means for reading and printing the system configuration data, a distinction should be drawn between raw configuration data and documentation. Raw configuration data often relates to a single system in a multi-system environment, and is mostly a cryptic list of variable names with their associated values, with little if any explanation of their meaning. An excellent example of raw configuration data may be found in Windows® Registry, which is a common configuration space for the Microsoft Windows® operating systems. Documentation on the other hand, is an organized collection of knowledge that, separately or in combination, represents the state or configuration of a system, teaches system operation, and aids in troubleshooting. Documentation is characteristically constructed of complete sentences in a human readable language, with appropriate punctuation, paragraph and optionally section separation. Typically, documentation also includes access tools such as a table of contents (TOC) or an index to further ease access to specific sections of the data Generally, documentation also attempts to explain the meaning behind the configuration parameters, optionally including the relationships between multiple systems and the meaning of several key parameters and their effect on system operation. Another optional feature of documentation is the ability to propose configurable parameter values for sample systems, or, in the case of the current invention, the system being documented. Optionally, drawings depicting the condition and interrelationship between the described system elements further enhance the documentation. It should also be noted that documentation is characterized primarily by its content and organization, and not by the medium on which it is presented. Thus documentation may be printed on paper, viewed or edited on a computer screen by the likes of a word processor or a World Wide Web browser, etc. Whether viewed on a computer screen or printed on paper, documentation helps preserve and manage the knowledge accumulated in an organization. It thus increases operational efficiency, helps educate new members, and aids in problem solving and design for future expansion.
There are currently several aids to grouping and displaying raw configuration data relating to computer systems. One such example is available from MessageWise® in Ottawa, Ontario, Canada. This system reads configuration data from several Microsoft Exchange servers, and places them in a database. Queries can then be placed against the data to find answers to specific question. However the MessageWise system does not provide a well-documented meaning of a textual and visual description, organized in an easily read document as described above.
Another configuration data collection tool named Emap and distributed by Microsoft® Corporation of Redmond Wash., provides a graphical representation of connections between exchange servers in an organization. Similarly, several products allow representing an organizational data or voice network by drawings representing the network structure. These types of products are exemplified in U.S. Pat. No. 5,926,463 to Ahern et al.
None of the above solutions provide documentation as described earlier in these specifications. All lack the effects and efficiency of teachings that a well organized textual and visual document provides, and thus do not provide the dept of knowledge preservation and knowledge management provided by such documents. It is clear therefore that there is an unfulfilled need in the industry to provide automated documentation generation, re-generation, and management tools for highly configurable systems.